3806 JACOBI AND ANDRE saimples were taken in the polar region above Ajaska (70°N) and in the equatorial region south of Hawaii (12° to 20°N). All three profiles of the Alaska series indicate a large nega- tive gradient in the 8- to 12-km layer, i.e., in the tropopause. Above this layer in all cases the concentration again increases to a maximum at about 15 to 16 km and then decreases with increasing altitude. For the interpretation of these results, two Rn*™profiles (A and B) are plotted in Figure 5. They were calculated for an exhalation rate of 1 atom/em® sec and the corresponding K profiles A and B are given on the right side of the figure. In the tropopause the gradient of profile B is similar to the observed slope during the Alaska series. If equilibrium conditions are assumed, this slope indicates a mean turbulent diffusion coefficient of 1 x 10‘ to 3 x 10‘ em?/sec in this air layer. In the Hawaii series the decrease in the tropopause region is less pronounced; it corresponds to an average diffusion coefficient of 5 x 10° to 1 x 10° cm’/sec. This high rate of turbulence is not consistent with the high verti- cal gradient of fission product and W™ activity in the lower equatorial stratosphere which was observed after the nuclear weapons tests in the equatorial stratosphere. It must be concluded that the high Rn™ content in the lower equatorial stratosphere is mainly due te upwarddirected convection, which may occur especially above continental areas, rather than to turbulent diffusion. The obseryed increase in concentration in the 12- to 15-km layer during the Alaska series can- not be explained by a steady-state equilibrium im an atmosphere which is horizontally isotropic. It may be explained either by horizontal advection of Rn*-enriched air in the lower stratosphere, which overlaps the tropopause layer, or bynonequilibrium conditions of the vertical exchange process at the sampling location. The first interpretation was given by Machta and Lucas [1962], who suggest that the Rn™ in the 15-km layer comes from the equatorial troposphere and enters the lower polar stratosphere through the tropopause gap. In this case the Rn™ in the 15-km layer above Alaska should have a more recent tropospheric history than that in the tropopause region below. Since the Rn™ content of this layer is about the same at polar and equatorial sampling sites, the transit time for the transport from south to north through the tropopause gap must be rather short, probably not exceeding a week. Another interpretation of the observed profile over Alaska is possible if a sudden temporary decrease of vertical mixing or convection is assumed. If, for instance, there is a sudden change of the K profile from type A to type B (Figure 5), the supply of fresh Rn™ from the troposphere to the stratosphere is interrupted. Because of the low turbulence rate in this layer, most of the old Rn™ in the lower stratosphere remains there until it decays. A fraction of the old Rn™ in the tropopause will diffuse downward into the troposphere or upward into the lower stratosphere. Since the mixing rate is higher in the tropopause than in the lower stratosphere, the Rn™ content decreases more rapidly in the tropopause than in the layer above, and the result is a temporary, slow increase in concentration with altitude above the tropopause. A check on possible contaminations of the air samples and further measurements are necessary to confirm these preliminary results. In any event they indicate the value of Rn™ as a tracer in the study of the exchange between troposphere and stratosphere. The main reason for its value is its rather short half-life of 3.8 days, which sets a time scale for the processes in- volved. Short-lived Rn’decay products. For the deeay products Po™(&). = 3.05 min), Pb™ (ti. = 26.8 min), and Bi™ (¢,. = 19.7 min), the theory predicts, as expected, radioactive equilibrium with Rn™ except in surface air. In the boundary layer the radioactive equilibrium is disturbed because of the deposition of decay products at the earth’s surface which results in a downward diffusion flux within the boundarylayer. Figure 6 shows the vertical profiles of Po, Pb™, and Bi“ (Po) in the boundary layer, which were calculated with the aid of the K profiles given in Figure 1. In all cases a lack of the three decay praducts with respect to Rn™ must be expected for steady-state conditions This deficiency decreases with increasing heigh above groundlevel. At a given height the Pb™, Rn™ and Bi™/Rn™ ratios are nearly equal bu are lower than the Po™/Rn™ratio. The height of the disturbed layer depends a: the mixing rate within the boundary layer. Fc